- Email: [email protected]
Low p53 and retinoblastoma protein expression in cervical intraepithelial neoplasia grade 3 lesions is associated with coexistent adenocarcinoma in situ
Human Pathology (2008) 39, 573–578
www.elsevier.com/locate/humpath
Original contribution
Low p53 and retinoblastoma protein expression in cervical intraepithelial neoplasia grade 3 lesions is associated with coexistent adenocarcinoma in situ Arnold-Jan Kruse MD, PhD a,d , Ivar Skaland MSc a , Ane Cecilie Munk MD a,b,c , Emiel Janssen PhD a , Einar Gudlaugsson MD a , Jan P.A. Baak MD, PhD, FRCPath, FIAC(Hon), Dr HonCausa(Antwerp) a,b,d,e,⁎ a
Department of Pathology, Stavanger University Hospital, PO Box 8100, 4068 Stavanger, Norway Department of Gynecology, Stavanger University Hospital, PO Box 8100, 4068 Stavanger, Norway c The Gade Institute, University of Bergen, 5000 Bergen, Norway d Department of Obstetrics and Gynaecology, Academic Medical Center Maastricht, 6202 Maastricht, The Netherlands e Free University, 1057 Amsterdam, The Netherlands b
Received 1 June 2007; revised 5 August 2007; accepted 6 August 2007
Keywords: CIN3s; AIS; p53; Retinoblastoma protein; Coexistence
Summary Clinically, it is important to distinguish cervical intraepithelial neoplasia grade 3 (CIN3) lesions with and without coexisting adenocarcinoma in situ (AIS), but endocervical curetting can be false negative. The frequency of high-risk human papillomavirus genotypes in CIN3 patients with and without AIS differs. CIN3 epithelial cell cycle regulator expression may reflect these differences and thereby indicate coexistent AIS. G1 pathway epithelial cell cycle regulators (pRb, p53, cyclin D, p16) and Ki-67 were analyzed by quantitative immunohistochemistry in CIN3s with and without AIS. Compared with the normal cervical squamous epithelium, the CIN3 epithelium in small punch biopsies showed strong diffuse p16 and Ki-67 expression. CIN3s with coexistent AIS had a significantly lower percentage of pRb (P = .03)- and p53 (P = .03)-positive nuclei in the lower half of the epithelium than CIN3s without coexistent AIS. None of the 10 cases with values of either pRb-positive nuclei 30% or greater or p53-positive nuclei 15% or greater in the lower half of the epithelium had coexistent AIS, contrasting 8 (24%) of the 33 cases with both low values of p53- and pRb-positive nuclei. Combined low p53 and pRb expression in the lower half of the epithelium in punch biopsies is associated with coexistent AIS in the cone. Despite the fact that the results of the current study are interesting and potentially clinically relevant, it should be emphasized that they must be confirmed according to Good Laboratory Practice in independent patient groups, preferably also in a prospective study. © 2008 Elsevier Inc. All rights reserved.
1. Introduction ⁎ Corresponding author. Department of Pathology, Stavanger University Hospital, PO Box 8100, 4068 Stavanger, Norway. E-mail address: [email protected] (J. P. A. Baak). 0046-8177/$ – see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.humpath.2007.08.014
Cervical carcinoma is the second most common cancer in women worldwide [1]. Chronic infection with high-risk human papillomavirus (hrHPV) is a necessary event in the
574 evolution of cervical carcinomas [2]. The incidence of squamous cell cancer has markedly declined in many developed countries, mainly because of cytologic screening [3]. However, even in some countries with wellorganized screening programs, the incidence of cervical adenocarcinomas has either remained the same or rather increased [3-5]. It has been suggested that this can be explained by the fact that glandular adenocarcinomas and their precursor lesions are often missed by conventional cytology, possibly because of their location higher in the cervical canal and therefore more awkward (or reduced) accessibility using the conventional brush sampling method [6]. Moreover, only 43% of patients with glandular lesions before conization of the cervix were detected by endocervical curettage (ECC) [7]. It could be clinically important to investigate the differences in etiology of cervical intraepithelial neoplasia grade 3 (CIN3) lesions with and without adenocarcinoma in situ (AIS). Indicators in CIN3 for possible coexistence of a glandular lesion can alert the gynecologist to perform an extensive cone biopsy high up in the endocervix. Such a high-risk AIS indicator also could serve as a safety net against false-negative ECCs. The frequency of hrHPV genotypes (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, and 68) differs for CIN3 patients with and without AIS [8]. In this study, it was shown that there was no significant difference in the frequency of HPV-16 or HPV-45 between CIN3 patients with or without AIS. However, the frequency of HPV-18 was significantly higher, and the frequency of hrHPV genotypes other than HPV-16, HPV-18, and HPV-45 was lower in cervical intraepithelial neoplasia lesions coexisting with AIS than in solitary cervical intraepithelial neoplasia lesions. This suggests that CIN3s, coexisting with highgrade glandular lesions, are etiologically different from squamous lesions without coexisting glandular lesions [8]. In cervical carcinogenesis, HPV is thought to inactivate the G1 pathway by inhibiting p53 and pRb proteins by E6 and E7 proteins [9,10]. p53 and pRb serve as potential markers for hrHPV E6 and E7 function [6]. However, the effectiveness of this inhibition depends on the HPV type [10-13]. Lesions with HPV-16 or 16-combination subtypes had a significantly lower percentage of pRb, p53, and cyclin D in the lower half of the epithelium than those with other HPV types [14]. Therefore, we evaluated the function of epithelial G1 cell cycle regulators in patients with CIN3s with or without coexisting AIS. Because the cervical epithelium is not homogeneous (cells move from the parabasal to the superficial layers over time), quantitative expression assessment was performed in a topographically specific way, that is, separately in the basal layer, the lower half, and the upper half of the epithelium. The number of immunopositive cells was carefully quantitated using image analysis equipment, rather than performing global estimations, to avoid lack of reproducibility caused by subjective estimations [15].
A. -J. Kruse et al.
2. Materials and methods 2.1. Patient selection, tissue samples, and revision of histology The Regional Medical Ethics Committee of Helse Vest, Norway, the Norwegian Data Inspection, and the Health Directorate of Norway approved the study. Punch biopsies from women who had histologic confirmed CIN3 without or with AIS were selected from the computerized pathology archives from our laboratory. Of these biopsies, 8 with and 35 without AIS were randomly selected for further investigation. (For each CIN3 with coexistent AIS, 4-5 CIN3s without AIS were selected of patients matched for diagnosis date and age.) The tissues had been routinely fixed in buffered 4% formaldehyde, embedded in paraffin, cut at 4 μm, and stained with hematoxylin and eosin (H&E). The histology of all samples was independently confirmed at review using the World Health Organization classification [3]. The revising pathologist was unaware of the original diagnosis and clinical data.
2.2. Immunohistochemistry Antigen retrieval and dilution of the antibodies were optimized before the study started. Paraffin sections of 4-μm thickness, adjacent to the H&E sections, were mounted onto silanized slides (DAKO, Glostrup, Denmark, S3002) and dried overnight at 37°C followed by 1 hour at 60°C. The sections were deparaffinized in xylene and rehydrated in a graded series of alcohol solutions. Antigen retrieval was performed by pressure cooking in 10 mmol/L Tris/1 mmol/L EDTA (pH 9.0) for 2 minutes at full pressure and cooling for 15 minutes. Immunostaining was performed using an autostainer (DAKO). TBS (tris-buffered saline) (S1968) was added at 0.05% and Tween 20 (pH 7.6) was the rinse buffer. Endogenous peroxidase activity was blocked by peroxidase blocking reagent S2001 (DAKO) for 10 minutes, and the sections were incubated with the monoclonal antibodies at the following dilutions: cyclin D (clone P2D11F11), 1:25 (Novocastra, Newcastle-upon-Tyne, UK); p16 (clone 6H12), 1:25 (Novocastra); Rb (clone 13A10), 1:25 (Novocastra); p53 (clone DO-7), 1:200 (DAKO); and Ki-67 (clone MIB-1), 1:100 (DAKO). DAKO antibody diluent S0809 was used, and the immune complex was visualized by peroxidase/DAB (ChemMate Envision Kit; DAKO, K 5007) with incubation of Envision/HRP rabbit mouse for 30 minutes, and DAB + chromogen for 10 minutes. The sections were counterstained with hematoxylin, dehydrated, and mounted. Controls for the immunostaining were performed using normal cervical tissue control sections and positive normal cell compartments within test sections. The section adjacent to the sections used for immunostaining was
p53, pRb and coexistence of AIS in CIN3
575
2.4. Statistical analysis Statistical analysis was performed by SPSS version 14 (SPSS Inc, Chicago, IL) for the percentage positive cells of the antibodies in the 3 layers of the epithelium using CIN3 with or without coexistent AIS as the independent variables (criterion standards). To investigate significant differences between the 2 groups, we used the Mann-Whitney U test. The level of significance was set to P b .05.
3. Results
Fig. 1 Diagram illustrating the different layers in which the measurements were performed.
cut and stained with H&E to ensure the presence of the same CIN3 in all test sections (“sandwich technique”).
2.3. Quantitation of expression patterns Evaluation of immunopositivity of the other antibodies was performed using quantitative image analysis equipment (QPRODIT version 6; Leica, Cambridge, UK). Cells were defined as either negative or positive. The localization of expression varied for the different antibodies (nuclear expression in case of p53, pRb, Ki-67, and cyclin D; combined nuclear/cytoplasmic for p16). The percentage of positive and negative cells was assessed by 2 independent observers in the basal cell layer, the lower half (excluding the basal cells) and the upper half of the epithelium (Fig. 1). The quantitation procedure resulted in continuous values for each biomarker analyzed, varying from 0% to 100%. In cases of interobserver disagreement of more than 10%, a repeated assessment was performed and a consensus obtained.
The median age of CIN3 patients with and without AIS did not differ (32 and 30 years; range, 21-82 and 27-82 years, respectively) (P = .20). Compared with normal squamous cervical epithelium next to the CIN3 lesions, diffuse homogeneous p16ink4a immunostaining and increased Ki67 expression was found in all CIN3s. There was an abrupt transition between the strongly diffuse pRb-positive normal epithelium and the much less positive/nearly negative CIN3s (particularly in the lower half of the epithelium), both in CIN3s without and with AIS. The intensity of diffuse p16 positivity, degree of Ki-67 positivity, or cyclin D positivity in CIN3s was not related to the coexistence of AIS. CIN3s with coexisting AIS had significantly lower expression of pRb (means, 9% versus 13%, P = .03) and p53 (means, 2% versus 9%, P = .03) in the lower half of the epithelium (Table 1, Fig. 2). Ten (23%) of the 43 CIN3 cases had values of either pRb-positive nuclei 30% or greater or p53-positive nuclei 15% or greater in the lower half of the epithelium. None of them had coexistent AIS (negative predictive value = 100%). In contrast, of the 33 CIN3 cases with both p53- and pRbpositive nuclei in the lower half of the epithelium below these values, 8 CIN3 cases had coexistent AIS (positive predictive value = 24%). Thus, a subgroup of 23% of all cases without coexistent AIS can be separated from the other
Table 1 Mean, SD, and the probability of no difference (P, Mann-Whitney U test) of all features measured between the CIN3 cases with AIS or not Immunoquantitative feature
CIN3 with AIS (n = 8)
CIN3 without AIS (n = 35)
Mean (SD)
Mean (SD)
% % % % % % % % % % % %
88.8 (31.8) 90.0 (24.5) 81.3 (29.5) 2.8 (3.6) 8.9 (9.0) 13.3 (14.7) 0.63 (1.8) 4.38 (7.3) 6.3 (13.8) 0.75 (1.8) 1.9 (2.6) 5.0 (10.4)
98 (29.4) 97.8 (6.7) 93.3 (20.0) 15.6 (32.3) 12.8 (7.5) 20.0 (18.5) 0.11 (0.33) 0.9 (6.2) 2.2 (10.1) 1.9 (3.5) 8.7 (3.4) 1.6 (3.2)
p16-positive nuclei, basal layer of the epithelium p16-positive nuclei, lower half of the epithelium p16-positive nuclei, upper half of the epithelium pRb-positive nuclei, basal layer of the epithelium pRb-positive nuclei, lower half of the epithelium pRb-positive nuclei, upper half of the epithelium cyclin D–positive nuclei, basal layer of the epithelium cyclin D–positive nuclei, lower half of the epithelium cyclin D–positive nuclei, upper half of the epithelium p53-positive nuclei, basal layer of the epithelium p53-positive nuclei, lower half of the epithelium p53-positive nuclei, upper half of the epithelium
P
.7 .7 .4 .6 .03 .3 .6 .9 .2 .3 .03 .6
576
A. -J. Kruse et al.
Fig. 2
An example of a CIN3 without (left) and with coexisting AIS (right).
p53, pRb and coexistence of AIS in CIN3
Fig. 3 Scatterplot of retinoblastoma protein (pRb) and p53 detection in the lower half of the epithelium of CIN3s with and without coexisting AIS. None of the 10 cases with values of either pRb-positive nuclei 30% or greater or p53-positive nuclei 15% or greater in the lower half of the epithelium had coexistent AIS, contrasting 8 (24%) of the 33 cases with both p53- and pRb-positive nuclei in the lower half of the epithelium below these values.
subgroup of 77% of all cases where coexistent AIS incidence is high (24%) (Fig. 3).
4. Discussion In cervical carcinogenesis, HPV is thought to inactivate the G1 pathway by inhibiting p53 and pRb proteins by E6 and E7 proteins [9,10]. The effectiveness of this inhibition depends on the HPV type [10-13]. Because the frequency of specific hrHPV genotypes differs significantly in CIN3 patients with and without AIS, immunoprofiling of pRb and p53 pathways in the CIN3 part may be predictive for coexistent AIS. The current results show that expression of p53 and pRb is significantly different for CIN3 cases with and without coexistent AIS. If one of these 2 cell cycle growth regulators is above a critical level, no coexistent AIS is found. It is important to note that AIS (which originates in the endocervix) is more difficult to detect both cytologically and colposcopically. In fact, false-negative Papanicolaou tests are more common in cases of AIS than of CIN3 [16,17]. The site of early glandular lesions does not usually permit them to be visualized [16]. Although the greater proportion of these false-negative Papanicolaou tests is attributable to a sampling error, a considerable proportion is caused by overlooking or misinterpretation by the cytologist. It has been suggested that this may be caused by the fact that
577 glandular cytologic abnormalities are considerably less common than squamous abnormalities, and therefore, most cytologists tend to be less comfortable recognizing and diagnosing them [18]. It is difficult to diagnose the AIS by colposcopy because this neoplasia can be hidden under normal metaplastic epithelium [19,20]. Furthermore, the squamocolumnar junction retracts to the endocervix with increasing age, and consequently, the colposcopy and biopsy procedure becomes less reliable for the diagnosis of AIS [21]. The current study shows that combined negativity/low positivity for p53 and pRb protein may be indicators in CIN3 for possible coexistence of a glandular lesion. If the results are confirmed in independent validation studies, they could be used as follows. Where the pathologist finds that p53 and pRb protein expression is both negative or very few nuclei are positive in the lower half of CIN3 epithelium of the punch biopsy, the gynecologist could be warned that this CIN3 patient has a higher-than-average risk of coexistent AIS. The gynecologist could then perform an extensive cone biopsy high up in the endocervix, even if the ECC is AIS negative. The pRb/p53 findings could thus act as a safety net for false-negative ECCs. The natural history of adenocarcinoma (in situ) of the cervix and its precursors is poorly understood, and literature regarding this subject is relatively sparse [22]. Zielinsky et al [6] showed that only a minority of cervical adenocarcinomas displayed absence of HPV DNA and immunostaining profiles, suggestive of an etiology independent of HPV. The CIN3 and AIS lesions in the present study showed a similar decrease of pRb-positive nuclei, in agreement with the important role of HPV (ie, HPV E7) in the carcinogenesis of adenocarcinomas and its precursors. This supports the common mechanistic model of development of these lesions. Moreover, as the distribution of HPV types differs significantly between the 2 precursor lesions, it seems that once dysplasia of either the squamous or glandular component is high grade, then the natural history is independent of the HPV type that caused the lesion to develop in the first place. In the current study, the p53 and pRb protein expression was determined in the differentiating epithelial cells. This is important because it takes into account the dynamic nature of the cervical epithelium. Epithelial cells are born in the parabasal layer and then mature and migrate to the surface (where they are desquamated). CIN is a dynamic process, and consequently, a particular microscopic image is just one snapshot (or “image frame”) in a long film of the dynamic changes over time in the life of a CIN3 under study [15]. The predictive information of p53 and pRb (the presence of a coexistent AIS) is contained in the lower half of the epithelium. This implies that this geography-specific analysis cannot be performed on cytologic specimens because cytologic specimens reflect only the exfoliated (top most) cells. The development of a diagnostic laboratory test with therapeutic consequences (ie, the application of p53 and pRb
578 expression patterns in the lower half of the epithelium of CIN3 lesions to predict the coexistence of AIS) requires the transition of certain well-defined developmental phases according to Good Laboratory Practice [23-26]. Although the current results are interesting, the value must be confirmed in independent patient sets before use as a routine method. We conclude that in this Good Laboratory Practice phase 1 study, low p53 and pRb expression in the lower half of the epithelium is a promising important predictor of coexistent AIS and must be confirmed in independent patient groups.
References [1] Ferlay J, Bray F, Pisani P, et al. Globocan 2000. Cancer incidence, mortality and prevalence worldwide. Lyon: IARC Press; 2001. [2] Bosch FX, Lorincz A, Munoz N, Meijer CJ, Shah KV. The causal relation between human papillomavirus and cervical cancer. J Clin Pathol 2002;55:244-65. [3] Tavassoli FA, Devilee P, editors. World Health Organization classification of tumours. Pathology and genetics of tumours of the breast and female genital organs. Lyon: IARC Press; 2003. p. 259-87. [4] Smith HO, Tiffany MF, Qualls CR, Key CR. The rising incidence of adenocarcinoma relative to squamous cell carcinoma of the uterine cervix in the United States—a 24-year population-based study. Gynecol Oncol 2000;78:97-105. [5] Sasieni P, Adams J. Changing rates of adenocarcinoma and adenosquamous carcinoma of the cervix in England. Lancet 2001; 357:1490-3. [6] Zielinsky GD, Snijder PJF, Rozendaal L, et al. The presence of highrisk HPV combined with specific p53 and p16ink4a expression patterns points to high-risk HPV as the main causative agent for adenocarcinoma in situ and adenocarcinoma of the cervix. J Pathol 2003;201:535-43. [7] Poynor EA, Barakat RR, Hoskins WJ. Management and follow-up of patients with adenocarcinoma in situ of the uterine cervix. Gynecol Oncol 1995;57:158-64. [8] Bekkers RL, Bulten J, Wiersma-van Tilburg A, et al. Coexisting highgrade glandular and squamous cervical lesions and human papillomavirus infections. Br J Cancer 2003;89:886-90. [9] Scheffner M, Werness BA, Huibregtse JM. The E6 oncoprotein encoded by human papillomavirus types 16 and 18 promotes the degradation of p53. Cell 1990;63:1129-36. [10] Dyson N, Howley PM, Munger K, Harlow E. The human papillomavirus 16 E7 oncoprotein is able to bind to the retinoblastoma gene product. Science 1989;23:934-9.
A. -J. Kruse et al. [11] Crook T, Tidy JA, Vousden KH. Degradation of p53 can be targeted by HPV E6 sequences distinct from those required for p53 binding and trans-activation. Cell 1991;67:547-56. [12] Heck DV, Yee CL, Howley PM, Munger K. Efficiency of binding the retinoblastoma protein correlates with the transforming capacity of the E7 oncoproteins of the human papillomaviruses. Proc Natl Acad Sci U S A 1992;89:4442-6. [13] Munger K, Werness BA, Dyson N, Phelps WC, Harlow E, Howley PM. Complex formation of human papillomavirus E7 proteins with the retinoblastoma tumor suppressor protein. EMBO J 1989;8:4099-105. [14] Baak JP, Kruse AJ, Garland SM, et al. Combined p53 and retinoblastoma protein detection identifies persistent and regressive cervical high-grade squamous intraepithelial lesions. Am J Surg Pathol 2005;29:1062-6. [15] Baak JP, Kruse AJ, Robboy SJ, Janssen EA, Van Diermen B, Skaland I. Dynamic behavioral interpretation of cervical intraepithelial neoplasia (CIN) by molecular biomarkers: a proposal. J Clin Pathol 2006;59:1017-28. [16] Mitchell H, Medley G, Gordon I, Giles G. Cervical cytology reported as negative risk of adenocarcinomas of uterine cervix: no strong evidence of benefit. Br J Cancer 1995;71:894-7. [17] Schwartz SM, Weiss NS. Increased incidence of adenocarcinoma of the cervix in young women in the United States. Am J Epidemiol 1986; 124:1045-7. [18] Wright TC, Kurman RJ, Ferenczy A. Precancerous lesions of the cervix. In: Kurman R, editor. Blaustein's pathology of the female genital tract. New York, USA: Springer-Verlag; 2002. p. 293-314. [19] Lickrish GM, Colgan TJ, Wright VC. Colposcopy of adenocarcinoma in situ and invasive adenocarcinoma of the cervix. Obstet Gynecol Clin North Am 1993;20:111-22. [20] Bertrand M, Lickrish GM, Colgan TJ. The anatomic distribution of cervical adenocarcinoma in situ: implication for the treatment. Am J Obstet Gynecol 1987;157:21-5. [21] Christopherson WMM, Nealon N, Gray LA. Noninvasive precursor lesions of adenocarcinoma and mixed adenosquamous carcinoma of the cervix uteri. Cancer 1979;44:975-83. [22] Ostor AG. Early invasive adenocarcinoma of the uterine cervix. Int J Gynecol Pathol 2000;19:29-38. [23] Dybkaer R, Jordal R, Jorgensen PJ, et al. A quality manual for the clinical laboratory—including the elements of a quality system. Proposed guidelines. Scan J Clin Lab Invest 1992;S52:1-27 (Draft 92-08-26). [24] Hall PA, Going JJ. Predicting the future: a critical appraisal of cancer prognosis studies. Histopathology 1999;35:489-94. [25] Makkink-Nombrado SV. Quantitative pathological, cell- and molecular biological markers in breast cancer. Towards more accurate prognostication and good laboratory practice in quantitative pathology. Thesis, Vrije Universiteit Amsterdam, The Netherlands, 1995, 4-9. [26] Baak JP. The framework of pathology: good laboratory practice by quantitative and molecular methods. J Pathol 2002;198:277-83.
www.elsevier.com/locate/humpath
Original contribution
Low p53 and retinoblastoma protein expression in cervical intraepithelial neoplasia grade 3 lesions is associated with coexistent adenocarcinoma in situ Arnold-Jan Kruse MD, PhD a,d , Ivar Skaland MSc a , Ane Cecilie Munk MD a,b,c , Emiel Janssen PhD a , Einar Gudlaugsson MD a , Jan P.A. Baak MD, PhD, FRCPath, FIAC(Hon), Dr HonCausa(Antwerp) a,b,d,e,⁎ a
Department of Pathology, Stavanger University Hospital, PO Box 8100, 4068 Stavanger, Norway Department of Gynecology, Stavanger University Hospital, PO Box 8100, 4068 Stavanger, Norway c The Gade Institute, University of Bergen, 5000 Bergen, Norway d Department of Obstetrics and Gynaecology, Academic Medical Center Maastricht, 6202 Maastricht, The Netherlands e Free University, 1057 Amsterdam, The Netherlands b
Received 1 June 2007; revised 5 August 2007; accepted 6 August 2007
Keywords: CIN3s; AIS; p53; Retinoblastoma protein; Coexistence
Summary Clinically, it is important to distinguish cervical intraepithelial neoplasia grade 3 (CIN3) lesions with and without coexisting adenocarcinoma in situ (AIS), but endocervical curetting can be false negative. The frequency of high-risk human papillomavirus genotypes in CIN3 patients with and without AIS differs. CIN3 epithelial cell cycle regulator expression may reflect these differences and thereby indicate coexistent AIS. G1 pathway epithelial cell cycle regulators (pRb, p53, cyclin D, p16) and Ki-67 were analyzed by quantitative immunohistochemistry in CIN3s with and without AIS. Compared with the normal cervical squamous epithelium, the CIN3 epithelium in small punch biopsies showed strong diffuse p16 and Ki-67 expression. CIN3s with coexistent AIS had a significantly lower percentage of pRb (P = .03)- and p53 (P = .03)-positive nuclei in the lower half of the epithelium than CIN3s without coexistent AIS. None of the 10 cases with values of either pRb-positive nuclei 30% or greater or p53-positive nuclei 15% or greater in the lower half of the epithelium had coexistent AIS, contrasting 8 (24%) of the 33 cases with both low values of p53- and pRb-positive nuclei. Combined low p53 and pRb expression in the lower half of the epithelium in punch biopsies is associated with coexistent AIS in the cone. Despite the fact that the results of the current study are interesting and potentially clinically relevant, it should be emphasized that they must be confirmed according to Good Laboratory Practice in independent patient groups, preferably also in a prospective study. © 2008 Elsevier Inc. All rights reserved.
1. Introduction ⁎ Corresponding author. Department of Pathology, Stavanger University Hospital, PO Box 8100, 4068 Stavanger, Norway. E-mail address: [email protected] (J. P. A. Baak). 0046-8177/$ – see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.humpath.2007.08.014
Cervical carcinoma is the second most common cancer in women worldwide [1]. Chronic infection with high-risk human papillomavirus (hrHPV) is a necessary event in the
574 evolution of cervical carcinomas [2]. The incidence of squamous cell cancer has markedly declined in many developed countries, mainly because of cytologic screening [3]. However, even in some countries with wellorganized screening programs, the incidence of cervical adenocarcinomas has either remained the same or rather increased [3-5]. It has been suggested that this can be explained by the fact that glandular adenocarcinomas and their precursor lesions are often missed by conventional cytology, possibly because of their location higher in the cervical canal and therefore more awkward (or reduced) accessibility using the conventional brush sampling method [6]. Moreover, only 43% of patients with glandular lesions before conization of the cervix were detected by endocervical curettage (ECC) [7]. It could be clinically important to investigate the differences in etiology of cervical intraepithelial neoplasia grade 3 (CIN3) lesions with and without adenocarcinoma in situ (AIS). Indicators in CIN3 for possible coexistence of a glandular lesion can alert the gynecologist to perform an extensive cone biopsy high up in the endocervix. Such a high-risk AIS indicator also could serve as a safety net against false-negative ECCs. The frequency of hrHPV genotypes (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, and 68) differs for CIN3 patients with and without AIS [8]. In this study, it was shown that there was no significant difference in the frequency of HPV-16 or HPV-45 between CIN3 patients with or without AIS. However, the frequency of HPV-18 was significantly higher, and the frequency of hrHPV genotypes other than HPV-16, HPV-18, and HPV-45 was lower in cervical intraepithelial neoplasia lesions coexisting with AIS than in solitary cervical intraepithelial neoplasia lesions. This suggests that CIN3s, coexisting with highgrade glandular lesions, are etiologically different from squamous lesions without coexisting glandular lesions [8]. In cervical carcinogenesis, HPV is thought to inactivate the G1 pathway by inhibiting p53 and pRb proteins by E6 and E7 proteins [9,10]. p53 and pRb serve as potential markers for hrHPV E6 and E7 function [6]. However, the effectiveness of this inhibition depends on the HPV type [10-13]. Lesions with HPV-16 or 16-combination subtypes had a significantly lower percentage of pRb, p53, and cyclin D in the lower half of the epithelium than those with other HPV types [14]. Therefore, we evaluated the function of epithelial G1 cell cycle regulators in patients with CIN3s with or without coexisting AIS. Because the cervical epithelium is not homogeneous (cells move from the parabasal to the superficial layers over time), quantitative expression assessment was performed in a topographically specific way, that is, separately in the basal layer, the lower half, and the upper half of the epithelium. The number of immunopositive cells was carefully quantitated using image analysis equipment, rather than performing global estimations, to avoid lack of reproducibility caused by subjective estimations [15].
A. -J. Kruse et al.
2. Materials and methods 2.1. Patient selection, tissue samples, and revision of histology The Regional Medical Ethics Committee of Helse Vest, Norway, the Norwegian Data Inspection, and the Health Directorate of Norway approved the study. Punch biopsies from women who had histologic confirmed CIN3 without or with AIS were selected from the computerized pathology archives from our laboratory. Of these biopsies, 8 with and 35 without AIS were randomly selected for further investigation. (For each CIN3 with coexistent AIS, 4-5 CIN3s without AIS were selected of patients matched for diagnosis date and age.) The tissues had been routinely fixed in buffered 4% formaldehyde, embedded in paraffin, cut at 4 μm, and stained with hematoxylin and eosin (H&E). The histology of all samples was independently confirmed at review using the World Health Organization classification [3]. The revising pathologist was unaware of the original diagnosis and clinical data.
2.2. Immunohistochemistry Antigen retrieval and dilution of the antibodies were optimized before the study started. Paraffin sections of 4-μm thickness, adjacent to the H&E sections, were mounted onto silanized slides (DAKO, Glostrup, Denmark, S3002) and dried overnight at 37°C followed by 1 hour at 60°C. The sections were deparaffinized in xylene and rehydrated in a graded series of alcohol solutions. Antigen retrieval was performed by pressure cooking in 10 mmol/L Tris/1 mmol/L EDTA (pH 9.0) for 2 minutes at full pressure and cooling for 15 minutes. Immunostaining was performed using an autostainer (DAKO). TBS (tris-buffered saline) (S1968) was added at 0.05% and Tween 20 (pH 7.6) was the rinse buffer. Endogenous peroxidase activity was blocked by peroxidase blocking reagent S2001 (DAKO) for 10 minutes, and the sections were incubated with the monoclonal antibodies at the following dilutions: cyclin D (clone P2D11F11), 1:25 (Novocastra, Newcastle-upon-Tyne, UK); p16 (clone 6H12), 1:25 (Novocastra); Rb (clone 13A10), 1:25 (Novocastra); p53 (clone DO-7), 1:200 (DAKO); and Ki-67 (clone MIB-1), 1:100 (DAKO). DAKO antibody diluent S0809 was used, and the immune complex was visualized by peroxidase/DAB (ChemMate Envision Kit; DAKO, K 5007) with incubation of Envision/HRP rabbit mouse for 30 minutes, and DAB + chromogen for 10 minutes. The sections were counterstained with hematoxylin, dehydrated, and mounted. Controls for the immunostaining were performed using normal cervical tissue control sections and positive normal cell compartments within test sections. The section adjacent to the sections used for immunostaining was
p53, pRb and coexistence of AIS in CIN3
575
2.4. Statistical analysis Statistical analysis was performed by SPSS version 14 (SPSS Inc, Chicago, IL) for the percentage positive cells of the antibodies in the 3 layers of the epithelium using CIN3 with or without coexistent AIS as the independent variables (criterion standards). To investigate significant differences between the 2 groups, we used the Mann-Whitney U test. The level of significance was set to P b .05.
3. Results
Fig. 1 Diagram illustrating the different layers in which the measurements were performed.
cut and stained with H&E to ensure the presence of the same CIN3 in all test sections (“sandwich technique”).
2.3. Quantitation of expression patterns Evaluation of immunopositivity of the other antibodies was performed using quantitative image analysis equipment (QPRODIT version 6; Leica, Cambridge, UK). Cells were defined as either negative or positive. The localization of expression varied for the different antibodies (nuclear expression in case of p53, pRb, Ki-67, and cyclin D; combined nuclear/cytoplasmic for p16). The percentage of positive and negative cells was assessed by 2 independent observers in the basal cell layer, the lower half (excluding the basal cells) and the upper half of the epithelium (Fig. 1). The quantitation procedure resulted in continuous values for each biomarker analyzed, varying from 0% to 100%. In cases of interobserver disagreement of more than 10%, a repeated assessment was performed and a consensus obtained.
The median age of CIN3 patients with and without AIS did not differ (32 and 30 years; range, 21-82 and 27-82 years, respectively) (P = .20). Compared with normal squamous cervical epithelium next to the CIN3 lesions, diffuse homogeneous p16ink4a immunostaining and increased Ki67 expression was found in all CIN3s. There was an abrupt transition between the strongly diffuse pRb-positive normal epithelium and the much less positive/nearly negative CIN3s (particularly in the lower half of the epithelium), both in CIN3s without and with AIS. The intensity of diffuse p16 positivity, degree of Ki-67 positivity, or cyclin D positivity in CIN3s was not related to the coexistence of AIS. CIN3s with coexisting AIS had significantly lower expression of pRb (means, 9% versus 13%, P = .03) and p53 (means, 2% versus 9%, P = .03) in the lower half of the epithelium (Table 1, Fig. 2). Ten (23%) of the 43 CIN3 cases had values of either pRb-positive nuclei 30% or greater or p53-positive nuclei 15% or greater in the lower half of the epithelium. None of them had coexistent AIS (negative predictive value = 100%). In contrast, of the 33 CIN3 cases with both p53- and pRbpositive nuclei in the lower half of the epithelium below these values, 8 CIN3 cases had coexistent AIS (positive predictive value = 24%). Thus, a subgroup of 23% of all cases without coexistent AIS can be separated from the other
Table 1 Mean, SD, and the probability of no difference (P, Mann-Whitney U test) of all features measured between the CIN3 cases with AIS or not Immunoquantitative feature
CIN3 with AIS (n = 8)
CIN3 without AIS (n = 35)
Mean (SD)
Mean (SD)
% % % % % % % % % % % %
88.8 (31.8) 90.0 (24.5) 81.3 (29.5) 2.8 (3.6) 8.9 (9.0) 13.3 (14.7) 0.63 (1.8) 4.38 (7.3) 6.3 (13.8) 0.75 (1.8) 1.9 (2.6) 5.0 (10.4)
98 (29.4) 97.8 (6.7) 93.3 (20.0) 15.6 (32.3) 12.8 (7.5) 20.0 (18.5) 0.11 (0.33) 0.9 (6.2) 2.2 (10.1) 1.9 (3.5) 8.7 (3.4) 1.6 (3.2)
p16-positive nuclei, basal layer of the epithelium p16-positive nuclei, lower half of the epithelium p16-positive nuclei, upper half of the epithelium pRb-positive nuclei, basal layer of the epithelium pRb-positive nuclei, lower half of the epithelium pRb-positive nuclei, upper half of the epithelium cyclin D–positive nuclei, basal layer of the epithelium cyclin D–positive nuclei, lower half of the epithelium cyclin D–positive nuclei, upper half of the epithelium p53-positive nuclei, basal layer of the epithelium p53-positive nuclei, lower half of the epithelium p53-positive nuclei, upper half of the epithelium
P
.7 .7 .4 .6 .03 .3 .6 .9 .2 .3 .03 .6
576
A. -J. Kruse et al.
Fig. 2
An example of a CIN3 without (left) and with coexisting AIS (right).
p53, pRb and coexistence of AIS in CIN3
Fig. 3 Scatterplot of retinoblastoma protein (pRb) and p53 detection in the lower half of the epithelium of CIN3s with and without coexisting AIS. None of the 10 cases with values of either pRb-positive nuclei 30% or greater or p53-positive nuclei 15% or greater in the lower half of the epithelium had coexistent AIS, contrasting 8 (24%) of the 33 cases with both p53- and pRb-positive nuclei in the lower half of the epithelium below these values.
subgroup of 77% of all cases where coexistent AIS incidence is high (24%) (Fig. 3).
4. Discussion In cervical carcinogenesis, HPV is thought to inactivate the G1 pathway by inhibiting p53 and pRb proteins by E6 and E7 proteins [9,10]. The effectiveness of this inhibition depends on the HPV type [10-13]. Because the frequency of specific hrHPV genotypes differs significantly in CIN3 patients with and without AIS, immunoprofiling of pRb and p53 pathways in the CIN3 part may be predictive for coexistent AIS. The current results show that expression of p53 and pRb is significantly different for CIN3 cases with and without coexistent AIS. If one of these 2 cell cycle growth regulators is above a critical level, no coexistent AIS is found. It is important to note that AIS (which originates in the endocervix) is more difficult to detect both cytologically and colposcopically. In fact, false-negative Papanicolaou tests are more common in cases of AIS than of CIN3 [16,17]. The site of early glandular lesions does not usually permit them to be visualized [16]. Although the greater proportion of these false-negative Papanicolaou tests is attributable to a sampling error, a considerable proportion is caused by overlooking or misinterpretation by the cytologist. It has been suggested that this may be caused by the fact that
577 glandular cytologic abnormalities are considerably less common than squamous abnormalities, and therefore, most cytologists tend to be less comfortable recognizing and diagnosing them [18]. It is difficult to diagnose the AIS by colposcopy because this neoplasia can be hidden under normal metaplastic epithelium [19,20]. Furthermore, the squamocolumnar junction retracts to the endocervix with increasing age, and consequently, the colposcopy and biopsy procedure becomes less reliable for the diagnosis of AIS [21]. The current study shows that combined negativity/low positivity for p53 and pRb protein may be indicators in CIN3 for possible coexistence of a glandular lesion. If the results are confirmed in independent validation studies, they could be used as follows. Where the pathologist finds that p53 and pRb protein expression is both negative or very few nuclei are positive in the lower half of CIN3 epithelium of the punch biopsy, the gynecologist could be warned that this CIN3 patient has a higher-than-average risk of coexistent AIS. The gynecologist could then perform an extensive cone biopsy high up in the endocervix, even if the ECC is AIS negative. The pRb/p53 findings could thus act as a safety net for false-negative ECCs. The natural history of adenocarcinoma (in situ) of the cervix and its precursors is poorly understood, and literature regarding this subject is relatively sparse [22]. Zielinsky et al [6] showed that only a minority of cervical adenocarcinomas displayed absence of HPV DNA and immunostaining profiles, suggestive of an etiology independent of HPV. The CIN3 and AIS lesions in the present study showed a similar decrease of pRb-positive nuclei, in agreement with the important role of HPV (ie, HPV E7) in the carcinogenesis of adenocarcinomas and its precursors. This supports the common mechanistic model of development of these lesions. Moreover, as the distribution of HPV types differs significantly between the 2 precursor lesions, it seems that once dysplasia of either the squamous or glandular component is high grade, then the natural history is independent of the HPV type that caused the lesion to develop in the first place. In the current study, the p53 and pRb protein expression was determined in the differentiating epithelial cells. This is important because it takes into account the dynamic nature of the cervical epithelium. Epithelial cells are born in the parabasal layer and then mature and migrate to the surface (where they are desquamated). CIN is a dynamic process, and consequently, a particular microscopic image is just one snapshot (or “image frame”) in a long film of the dynamic changes over time in the life of a CIN3 under study [15]. The predictive information of p53 and pRb (the presence of a coexistent AIS) is contained in the lower half of the epithelium. This implies that this geography-specific analysis cannot be performed on cytologic specimens because cytologic specimens reflect only the exfoliated (top most) cells. The development of a diagnostic laboratory test with therapeutic consequences (ie, the application of p53 and pRb
578 expression patterns in the lower half of the epithelium of CIN3 lesions to predict the coexistence of AIS) requires the transition of certain well-defined developmental phases according to Good Laboratory Practice [23-26]. Although the current results are interesting, the value must be confirmed in independent patient sets before use as a routine method. We conclude that in this Good Laboratory Practice phase 1 study, low p53 and pRb expression in the lower half of the epithelium is a promising important predictor of coexistent AIS and must be confirmed in independent patient groups.
References [1] Ferlay J, Bray F, Pisani P, et al. Globocan 2000. Cancer incidence, mortality and prevalence worldwide. Lyon: IARC Press; 2001. [2] Bosch FX, Lorincz A, Munoz N, Meijer CJ, Shah KV. The causal relation between human papillomavirus and cervical cancer. J Clin Pathol 2002;55:244-65. [3] Tavassoli FA, Devilee P, editors. World Health Organization classification of tumours. Pathology and genetics of tumours of the breast and female genital organs. Lyon: IARC Press; 2003. p. 259-87. [4] Smith HO, Tiffany MF, Qualls CR, Key CR. The rising incidence of adenocarcinoma relative to squamous cell carcinoma of the uterine cervix in the United States—a 24-year population-based study. Gynecol Oncol 2000;78:97-105. [5] Sasieni P, Adams J. Changing rates of adenocarcinoma and adenosquamous carcinoma of the cervix in England. Lancet 2001; 357:1490-3. [6] Zielinsky GD, Snijder PJF, Rozendaal L, et al. The presence of highrisk HPV combined with specific p53 and p16ink4a expression patterns points to high-risk HPV as the main causative agent for adenocarcinoma in situ and adenocarcinoma of the cervix. J Pathol 2003;201:535-43. [7] Poynor EA, Barakat RR, Hoskins WJ. Management and follow-up of patients with adenocarcinoma in situ of the uterine cervix. Gynecol Oncol 1995;57:158-64. [8] Bekkers RL, Bulten J, Wiersma-van Tilburg A, et al. Coexisting highgrade glandular and squamous cervical lesions and human papillomavirus infections. Br J Cancer 2003;89:886-90. [9] Scheffner M, Werness BA, Huibregtse JM. The E6 oncoprotein encoded by human papillomavirus types 16 and 18 promotes the degradation of p53. Cell 1990;63:1129-36. [10] Dyson N, Howley PM, Munger K, Harlow E. The human papillomavirus 16 E7 oncoprotein is able to bind to the retinoblastoma gene product. Science 1989;23:934-9.
A. -J. Kruse et al. [11] Crook T, Tidy JA, Vousden KH. Degradation of p53 can be targeted by HPV E6 sequences distinct from those required for p53 binding and trans-activation. Cell 1991;67:547-56. [12] Heck DV, Yee CL, Howley PM, Munger K. Efficiency of binding the retinoblastoma protein correlates with the transforming capacity of the E7 oncoproteins of the human papillomaviruses. Proc Natl Acad Sci U S A 1992;89:4442-6. [13] Munger K, Werness BA, Dyson N, Phelps WC, Harlow E, Howley PM. Complex formation of human papillomavirus E7 proteins with the retinoblastoma tumor suppressor protein. EMBO J 1989;8:4099-105. [14] Baak JP, Kruse AJ, Garland SM, et al. Combined p53 and retinoblastoma protein detection identifies persistent and regressive cervical high-grade squamous intraepithelial lesions. Am J Surg Pathol 2005;29:1062-6. [15] Baak JP, Kruse AJ, Robboy SJ, Janssen EA, Van Diermen B, Skaland I. Dynamic behavioral interpretation of cervical intraepithelial neoplasia (CIN) by molecular biomarkers: a proposal. J Clin Pathol 2006;59:1017-28. [16] Mitchell H, Medley G, Gordon I, Giles G. Cervical cytology reported as negative risk of adenocarcinomas of uterine cervix: no strong evidence of benefit. Br J Cancer 1995;71:894-7. [17] Schwartz SM, Weiss NS. Increased incidence of adenocarcinoma of the cervix in young women in the United States. Am J Epidemiol 1986; 124:1045-7. [18] Wright TC, Kurman RJ, Ferenczy A. Precancerous lesions of the cervix. In: Kurman R, editor. Blaustein's pathology of the female genital tract. New York, USA: Springer-Verlag; 2002. p. 293-314. [19] Lickrish GM, Colgan TJ, Wright VC. Colposcopy of adenocarcinoma in situ and invasive adenocarcinoma of the cervix. Obstet Gynecol Clin North Am 1993;20:111-22. [20] Bertrand M, Lickrish GM, Colgan TJ. The anatomic distribution of cervical adenocarcinoma in situ: implication for the treatment. Am J Obstet Gynecol 1987;157:21-5. [21] Christopherson WMM, Nealon N, Gray LA. Noninvasive precursor lesions of adenocarcinoma and mixed adenosquamous carcinoma of the cervix uteri. Cancer 1979;44:975-83. [22] Ostor AG. Early invasive adenocarcinoma of the uterine cervix. Int J Gynecol Pathol 2000;19:29-38. [23] Dybkaer R, Jordal R, Jorgensen PJ, et al. A quality manual for the clinical laboratory—including the elements of a quality system. Proposed guidelines. Scan J Clin Lab Invest 1992;S52:1-27 (Draft 92-08-26). [24] Hall PA, Going JJ. Predicting the future: a critical appraisal of cancer prognosis studies. Histopathology 1999;35:489-94. [25] Makkink-Nombrado SV. Quantitative pathological, cell- and molecular biological markers in breast cancer. Towards more accurate prognostication and good laboratory practice in quantitative pathology. Thesis, Vrije Universiteit Amsterdam, The Netherlands, 1995, 4-9. [26] Baak JP. The framework of pathology: good laboratory practice by quantitative and molecular methods. J Pathol 2002;198:277-83.